A method and device for power grid frequency regulation and a storage medium

Abstract

The application discloses a power grid frequency regulation method and device, and a storage medium. The method comprises the following steps: calculating a secondary frequency modulation optimization objective function according to a power supply cost coefficient, a frequency modulation mileage cost and an optimization weight coefficient of a frequency deviation; constructing a frequency response model of a power grid system according to load and power supply frequency modulation standby data, power supply model information and load data, generating state variables, control variables and disturbance variables, and constructing a system prediction state model; establishing a system frequency optimization control model with the minimum secondary frequency modulation optimization objective function and the system prediction state model as a constraint condition; and solving the system frequency optimization control model by using a distributed optimization solving algorithm, so that the optimization objective function meets a preset requirement, the optimal decision variable is obtained as a frequency regulation result, the accuracy of power grid frequency control is improved, and the influence of new energy and load prediction errors and random disturbances on the system frequency regulation performance is reduced.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, apparatus and storage medium for power grid frequency regulation. Background Technology

[0002] Currently, the scale of new energy sources and their proportion in the power grid are constantly increasing. Direct grid connection of new energy sources will bring huge uncertainties to the safe operation of the power system. In addition, flexible resources such as photovoltaic, energy storage equipment, and electric vehicles controlled by power electronic equipment can quickly respond to the grid's control commands and provide auxiliary services such as frequency regulation and virtual inertia.

[0003] To address the operational and cost characteristics of large-scale, multi-type distributed resources, existing power system frequency control algorithms often employ centralized methods, placing high demands on the system's communication network and the computational capabilities of the centralized controller. This approach is generally unsuitable for real-time control of power systems with massive distributed resources. While existing distributed optimization algorithms only require information exchange between adjacent power sources, each optimization sub-problem unit still needs to provide the original variables, dual variables, and other auxiliary variables for algorithm calculation, resulting in a significant computational burden, complex communication topologies, and low computational performance. Furthermore, in existing distributed optimization algorithms, computational errors and random disturbances have a significant impact on system frequency regulation performance, leading to substantial system uncertainty. Summary of the Invention

[0004] This invention provides a method, apparatus, and storage medium for power grid frequency regulation, in order to improve the accuracy of power grid frequency control and reduce the impact of new energy and load forecasting errors and random disturbances on the system frequency regulation performance.

[0005] To improve the accuracy of power grid frequency control, this invention provides a method for power grid frequency regulation, including: acquiring load and power supply frequency regulation reserve data, power supply model information, load data, and power supply cost coefficient of the power grid system;

[0006] Based on the power cost coefficient and the reference output active power of the distributed resource cluster, a first cost function for the active power regulation of the distributed resource cluster is calculated; based on the first cost function, the frequency regulation mileage cost, and the optimization weight coefficient of the frequency deviation, the sum of the second cost functions from the current period to the future preset period is calculated as the optimization objective function of the secondary frequency regulation.

[0007] Based on the load and power supply frequency regulation reserve data, the power supply model information, and the load data, a frequency response model of the power grid system is constructed, state variables, control variables, and disturbance variables are generated, and a system predictive state model is constructed.

[0008] A system frequency optimization control model is established with the objective function of minimizing the secondary frequency regulation as the constraint condition, and with the output constraints of the system predictive state model, the synchronous machine, the renewable energy unit and the energy storage system as the constraints.

[0009] The system frequency optimization control model is solved by a distributed optimization algorithm, so that the optimization objective function meets the preset requirements and the optimal decision variable is obtained as the frequency regulation result.

[0010] As a preferred embodiment, the power grid frequency regulation method proposed in this invention uses the sum of the second cost functions from the current cycle to a future preset cycle as the optimization objective function for secondary frequency regulation, and performs frequency optimization control by minimizing the optimization objective function of secondary frequency regulation, thereby reducing the system operating cost during the secondary frequency regulation stage. Furthermore, by constructing a system prediction state model using state variables, control variables, and disturbance variables, errors in new energy prediction data are reduced, ensuring the robustness of frequency control, improving the accuracy of power grid frequency control, and reducing the impact of prediction errors and random disturbances on the system frequency regulation performance. In addition, this invention uses a distributed optimization solution algorithm to solve the system frequency optimization control model, reducing the computational complexity and communication burden of real-time frequency control.

[0011] As a preferred embodiment, based on the load and power supply frequency regulation reserve data, the power supply model information, and the load data, a frequency response model of the power grid system is constructed, generating state variables, control variables, and disturbance variables, and a system predictive state model is constructed, specifically as follows:

[0012] Based on load and power supply frequency regulation reserve data and power supply model information, the upper and lower limits of the output of various power sources in the power grid system are generated; based on the upper and lower limits of the output, the active power output value of each of the power sources is calculated, and the active power output value of all the power sources is used as the secondary frequency regulation model of the distributed resource cluster.

[0013] Based on the load data and the secondary frequency regulation model, a frequency response model of the power grid system is constructed;

[0014] Based on the frequency response model and the second frequency modulation model, state variables, control variables, and disturbance variables are generated, and a system prediction state model is constructed.

[0015] As a preferred embodiment, this invention further eliminates the frequency deviation of the power system through real-time secondary frequency regulation of various controllable power sources, improves the accuracy of power grid frequency control, and reduces the error in new energy prediction data by constructing a system prediction state model through state variables, control variables, and disturbance variables, ensuring the robustness of frequency control, improving the accuracy of power grid frequency control, and reducing the impact of prediction errors and random disturbances on the system frequency regulation performance.

[0016] As a preferred embodiment, based on load and power supply frequency regulation reserve data and power supply model information, upper and lower limits for the output of various power sources in the power grid system are generated. Based on these upper and lower limits, the active power output values ​​of each power source are calculated. The active power output values ​​of all power sources are then used as the secondary frequency regulation model for the distributed resource cluster. Specifically:

[0017] Based on load and power frequency regulation reserve data and power model information, upper and lower limits of control commands for synchronous machines, new energy units and energy storage devices in the power grid system are generated respectively.

[0018] Based on the dynamic adjustment time constant, control command, and upper and lower limits of the control command for various power sources, the active power output values ​​of the synchronous machine, new energy units, and energy storage devices in the power grid system are generated respectively.

[0019] The upper and lower limits of the control commands of the energy storage device are calculated based on the charging and discharging efficiency of the energy storage device and the lower and upper limits of the controllable adjustable amount of the energy storage device.

[0020] The active power output values ​​of synchronous machines, new energy generating units, and energy storage devices in the power grid system are used as the secondary frequency regulation model of the distributed resource cluster.

[0021] As a preferred embodiment, the present invention uses the active power output values ​​of the synchronous machine, new energy generating units, and energy storage devices as the secondary frequency regulation model of the distributed resource cluster. The real-time secondary frequency regulation of the synchronous machine, new energy generating units, and energy storage devices further eliminates the frequency deviation of the power system and improves the accuracy of power grid frequency control.

[0022] As a preferred embodiment, a frequency response model of the power grid system is constructed based on the load data and the secondary frequency regulation model, specifically as follows:

[0023] Based on the active power output values ​​of synchronous machines, new energy generating units, and energy storage devices, load data, and the oscillation equation of the power grid system, a frequency response model of the power grid system is constructed:

[0024]

[0025] Where G is the set of synchronous machines, ΔP i SG R is the active power output value of synchronous machine i, and R is the set of new energy generating units. Here, E represents the active power output of the new energy unit j, and E represents the collection of energy storage devices. Where M is the active power output of energy storage device j, D is the overall system inertia, and P is the overall system damping coefficient. L Here, Δf represents the system load data, ΔP represents the frequency deviation of the power grid system, and ΔP represents the frequency deviation of the power grid system. d It is the disturbance quantity of the system's active power change.

[0026] As a preferred embodiment, the frequency response model of the power grid system constructed in this invention comprehensively considers the relationship between the active power output values ​​of different types of power sources, such as synchronous machines, new energy generating units, and energy storage devices, and the frequency deviation of the power grid system. The frequency deviation of the power system is further eliminated through real-time secondary frequency regulation by synchronous machines, new energy generating units, and energy storage devices, thereby improving the accuracy of power grid frequency control.

[0027] As a preferred embodiment, based on the frequency response model and the second-order frequency modulation model, state variables, control variables, and disturbance variables are generated, and a system prediction state model is constructed, specifically as follows:

[0028] Based on the frequency deviation of the power grid system, the active power output of synchronous machine i, the active power output of new energy unit j, and the active power output of energy storage device j, state variables are constructed; based on the control commands of the synchronous machine, new energy unit, and energy storage device, control variables are constructed; based on the disturbance amount of the system's active power change, disturbance variables are constructed.

[0029] Based on the state variables, control variables, and disturbance variables, construct the discrete state-space model of the system: x(k+1)=Ax(k)+Bu(k)+D d d(k);

[0030] Where x(k), u(k), and d(k) are the system's state variable, control variable, and disturbance variable, respectively; A, B, and D d These are the system's state matrix, input matrix, and disturbance matrix, respectively.

[0031] The disturbance matrix is ​​expanded into a prediction model for N time periods as the system prediction state model:

[0032] X = S x x(0)+S u U+S d D d ;

[0033] Where X is the state variable matrix, x(0) is the first element of the state variable matrix; U is the control command matrix, D d Let X, U, and D be the perturbation variable matrix. d Both are extensions of the original system state vector at N time points; S x S u and S d Let be the extended matrices of the system's state matrix, input matrix, and disturbance matrix at N time points.

[0034] As a preferred embodiment, this invention establishes a discrete state-space model for secondary frequency regulation, including state variables, control variables, and disturbance variables, as the MPC state-space model for secondary frequency regulation. The MPC state-space model of this system extends the original system model into a predictive model for N time points, describing the dynamic frequency response of the system at N time points. Through real-time secondary frequency regulation by synchronous machines, new energy generating units, and energy storage devices, the frequency deviation of the power system is further eliminated, and the accuracy of power grid frequency control is improved.

[0035] As a preferred option, a system frequency optimization control model is established with the objective function of minimizing secondary frequency regulation as the constraint condition, and the output constraints of the system predictive state model, the synchronous machine, the renewable energy unit, and the energy storage system. Specifically:

[0036]

[0037] stX=S x x(0)+S u U+S d D d

[0038]

[0039]

[0040]

[0041] Among them, J SFC Let be the objective function for optimizing the second frequency modulation. It is the input control command of the synchronizer i. and These are the upper and lower limits of the control commands for synchronizer i. This is the control command for the new energy unit J. and These are the upper and lower limits of the control commands for the new energy generating unit j. These are the control commands for energy storage device j. and These represent the upper and lower limits of the control commands for energy storage device j, respectively. The upper and lower bounds of the output constraint are subtracted from the output estimate of the power supply in the primary frequency regulation, where P is the set of power supplies (new energy units and energy storage devices) at the inverter interface.

[0042] As a preferred embodiment, the system frequency optimization control model of the present invention reduces the operating cost of frequency regulation ancillary services while meeting frequency security constraints, and further eliminates the frequency deviation of the power system through real-time secondary frequency regulation by synchronous machines, new energy units and energy storage devices, thereby improving the accuracy of power grid frequency control.

[0043] As a preferred option, the upper and lower bounds of the output constraint minus the output estimate of the power supply in the primary frequency regulation is calculated from the virtual inertia coefficient and the droop control parameters, specifically:

[0044]

[0045] in, and denoted as the virtual inertia coefficient and droop control parameter of the distributed resource cluster, respectively, and f(k) and Δf(k) are the frequency change rate and frequency deviation measurement values ​​at time k, respectively.

[0046] As a preferred embodiment, the frequency regulation method of the present invention can reduce the operating cost of frequency regulation auxiliary services while meeting frequency security constraints, and coordinate the reserve and power output relationship of new energy units and energy storage devices in the inertial response and secondary frequency regulation stages.

[0047] As a preferred approach, the system frequency optimization control model is solved using a distributed optimization algorithm, so that the optimization objective function meets the preset requirements, and the optimal decision variables are obtained as the frequency regulation result. Specifically:

[0048] The system frequency optimization control model is transformed into:

[0049]

[0050] stA eq,i X i =b eq,i

[0051] g i (X i )≤0;

[0052] The optimization problem is divided into m subproblems, J i Let X be the cost function of subproblem i. i Let X be the decision variable for subproblem i. i Includes state variables, control commands, and disturbance variables; A eq,i b eq,i These correspond to the linear equality constraints and constant control variables in the system frequency optimization control model, respectively; g i (X i ) represents the expression for the inequality constraints, corresponding to the constraints on all power supply outputs in the system frequency optimization control model.

[0053] Based on the preset iteration step size constant, update the intermediate variable of power supply i in the nth iteration. Control all power sources to interact with their adjacent power sources one by one with their updated intermediate variables; n is the number of iterations;

[0054] According to the preset consensus update rule of distributed optimization, the cost function in the (n+1)th iteration is calculated, and it is determined whether the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition. If so, the current decision variable is used as the frequency regulation result.

[0055] If not, the iteration count is incremented by one until the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition.

[0056] As a preferred embodiment, this invention solves the system frequency optimization control model by using a distributed optimization algorithm based on neurodynamics, which reduces the computational burden of real-time control and simplifies the communication topology between various power sources. Compared with existing centralized and distributed control algorithms, it can exhibit better computational efficiency and accuracy.

[0057] Accordingly, the present invention also provides a device for power grid frequency regulation, comprising: an acquisition module, an objective function calculation module, a state prediction module, an optimization control module, and a regulation result output module;

[0058] The acquisition module is used to acquire load and power frequency regulation reserve data, power model information, load data and power cost coefficient of the power grid system.

[0059] The objective function calculation module is used to calculate the first cost function of the active power regulation of the distributed resource cluster based on the power cost coefficient and the reference output active power of the distributed resource cluster; and to calculate the sum of the second cost functions from the current period to the future preset period as the optimization objective function of the secondary frequency regulation based on the first cost function, the frequency regulation mileage cost and the optimization weight coefficient of the frequency deviation.

[0060] The state prediction module is used to construct a frequency response model of the power grid system based on the load and power supply frequency regulation reserve data, the power supply model information and the load data, generate state variables, control variables and disturbance variables, and construct a system predicted state model.

[0061] The optimization control module is used to establish a system frequency optimization control model with the objective function of secondary frequency regulation being minimized and with constraints such as the system predictive state model, the output constraints of the synchronous machine, the output constraints of the renewable energy unit, and the output constraints of the energy storage system.

[0062] The regulation result output module is used to solve the system frequency optimization control model through a distributed optimization solution algorithm, so that the optimization objective function reaches the preset requirements and the optimal decision variable is obtained as the frequency regulation result.

[0063] As a preferred embodiment, the objective function calculation module of this invention uses the sum of the second cost functions from the current period to a future preset period as the optimization objective function for secondary frequency regulation. The optimization control module performs frequency optimization control by minimizing the optimization objective function of secondary frequency regulation, thereby reducing the system operating cost during the secondary frequency regulation stage. The state prediction module constructs a system prediction state model using state variables, control variables, and disturbance variables, reducing errors in new energy prediction data, ensuring the robustness of frequency control, improving the accuracy of grid frequency control, and reducing the impact of prediction errors and random disturbances on the system frequency regulation performance. Furthermore, the control result output module of this invention solves the system frequency optimization control model using a distributed optimization algorithm, reducing the computational complexity and communication burden of real-time frequency control.

[0064] Accordingly, the present invention also provides a computer-readable storage medium comprising a stored computer program; wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform a power grid frequency regulation method as described in the present invention. Attached Figure Description

[0065] Figure 1 This is a flowchart illustrating one embodiment of a power grid frequency regulation method provided by the present invention;

[0066] Figure 2 This is a schematic diagram of one embodiment of a power grid frequency regulation device provided by the present invention. Detailed Implementation

[0067] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] Example 1

[0069] Please refer to Figure 1 The present invention provides a method for power grid frequency regulation, comprising steps S101-S105:

[0070] Step S101: Obtain load and power frequency regulation reserve data, power model information, load data and power cost coefficient of the power grid system;

[0071] In this embodiment, the power grid frequency regulation method provided by the present invention is used for real-time frequency control of the power system. Before real-time control, load and power frequency regulation reserve data are obtained intraday or in real time. Preferably, the real-time load and power frequency regulation reserve data is updated every 15 minutes.

[0072] In this embodiment, the load data refers to the system's load level; the power supply frequency regulation reserve data includes the upper and lower limits of the output for each type of power supply; and the power supply cost factor is the primary and secondary cost factor a. i and b i And the FM mileage cost coefficient r i The power model information includes the dynamic adjustment time constant, upper and lower limits of control commands, and charging and discharging efficiency of energy storage devices for all types of power sources.

[0073] Step S102: Calculate the first cost function of the active power regulation of the distributed resource cluster based on the power cost coefficient and the reference output active power of the distributed resource cluster; calculate the sum of the second cost functions from the current period to the future preset period based on the first cost function, the frequency regulation mileage cost and the optimization weight coefficient of the frequency deviation, as the optimization objective function of the secondary frequency regulation.

[0074] In this embodiment, the active power adjustment ΔP of the distributed resource cluster i The first cost function of (k) is expressed as:

[0075] ΔC i (ΔP i (k))=a i (ΔP i (k)) 2 +(2a i P i (k)+b i )ΔP i (k);

[0076] Among them, P i (k) represents the reference output active power ΔP in the scheduling plan of distributed resource cluster i at time k. i (k) represents the active power adjustment of resource cluster i at time k, a i and b i This represents the cost coefficient for primary and secondary scheduling of the distributed resource cluster.

[0077] To avoid damage to the speed controller caused by frequent adjustments, let r... i Let Δu be the FM mileage cost factor. i (k) and Δu i (k-1) represents the secondary frequency modulation command for two consecutive control cycles, defining the frequency modulation mileage cost ΔL.i as follows:

[0078] ΔL i (Δu i (k))=r i |Δu i (k)-Δu i (k-1)|;

[0079] The sum of the second cost functions from the current period k to the next N periods is used as the optimization objective function J for the second frequency modulation. SFC :

[0080]

[0081] Where G is the set of all synchronous machines i in the system, P is the set of all power electronic interface asynchronous power supplies j in the system, and q f q is the optimized weighting coefficient for the frequency deviation term Δf(k). f The q value determines the system's ability to adjust for frequency deviation in secondary frequency modulation. f The larger the value, the stronger the frequency deviation adjustment effect.

[0082] Step S103: Based on the load and power supply frequency regulation reserve data, the power supply model information and the load data, construct a frequency response model of the power grid system, generate state variables, control variables and disturbance variables, and construct a system predictive state model;

[0083] In this embodiment, based on the load and power supply frequency regulation reserve data, the power supply model information, and the load data, a frequency response model of the power grid system is constructed, generating state variables, control variables, and disturbance variables, and a system predictive state model is constructed, specifically as follows:

[0084] Based on load and power supply frequency regulation reserve data and power supply model information, the upper and lower limits of the output of various power sources in the power grid system are generated; based on the upper and lower limits of the output, the active power output value of each of the power sources is calculated, and the active power output value of all the power sources is used as the secondary frequency regulation model of the distributed resource cluster.

[0085] Based on the load data and the secondary frequency regulation model, a frequency response model of the power grid system is constructed;

[0086] Based on the frequency response model and the second frequency modulation model, state variables, control variables, and disturbance variables are generated, and a system prediction state model is constructed.

[0087] In this embodiment, based on the upper and lower limits of the output, the active power output values ​​of various types of power sources are calculated, and the active power output values ​​of all power sources are used as the secondary frequency regulation model of the distributed resource cluster, specifically:

[0088] Establish a secondary frequency modulation model for distributed resource clusters:

[0089]

[0090]

[0091]

[0092]

[0093]

[0094] Wherein, ΔP i SG This is the active power output of synchronous machine i. It is the dynamic adjustment time constant of the synchronous machine. It is the input control command of the synchronizer i. and These are the upper and lower limits of the control commands for synchronizer i. It is the active power output of the new energy unit j. It is the dynamic adjustment time constant of the new energy unit. This is the control command for the new energy unit J. and These are the upper and lower limits of the control commands for the new energy generating unit j. It is the active power output of energy storage device j. It is the dynamic adjustment time constant of the energy storage device. These are the control commands for energy storage device j. and These are the upper and lower limits of the control command for energy storage device j, respectively, and E j (k) and E j (k+1) represents the energy levels of the energy storage device at time k and time k+1, respectively. and The charging and discharging efficiency of energy storage devices. These are the lower and upper limits of the controllable adjustable power of the energy storage device j. and E j These represent the upper and lower limits of the electrical energy level of energy storage device j, respectively. At time k, the lower and upper limits of the actual power regulation of energy storage device j are respectively determined by... and This indicates that the actual adjustment upper limit is taken as the upper limit of the controllable adjustable power. The minimum value of the remaining charge allowable output power, where the remaining charge allowable output power is determined by the remaining charge E at time k. j (k)-E j Multiply by discharge efficiency Divide by the unit time Δt to obtain the result; similarly, the lower limit of the actual regulating power is taken as the lower limit of the controllable regulating power. The maximum value of the allowable charging power at time k, where the allowable charging power is increased by the amount of electricity at time k. Divide by charging efficiency The values ​​are obtained from the unit time Δt. All upper and lower limits of the power output are determined by the intraday optimization reserve, and the results of the intraday optimization scheduling are known during the real-time control phase.

[0095] In this embodiment, a frequency response model of the power grid system is constructed based on the load data and the secondary frequency regulation model, specifically as follows:

[0096] Based on the active power output values ​​of synchronous machines, new energy generating units, and energy storage devices, load data, and the oscillation equation of the power grid system, a frequency response model of the power grid system is constructed:

[0097]

[0098] Where G is the set of synchronous machines, ΔP i SG R is the active power output value of synchronous machine i, and R is the set of new energy generating units. Here, E represents the active power output of the new energy unit j, and E represents the collection of energy storage devices. Where M is the active power output of energy storage device j, D is the overall system inertia, and P is the overall system damping coefficient. L Here, Δf represents the system load data, ΔP represents the frequency deviation of the power grid system, and ΔP represents the frequency deviation of the power grid system. d It is the disturbance quantity of the system's active power change.

[0099] In this embodiment, based on the frequency response model and the second-order frequency modulation model, state variables, control variables, and disturbance variables are generated, and a system prediction state model is constructed, specifically as follows:

[0100] Based on the frequency deviation of the power grid system, the active power output of synchronous machine i, the active power output of new energy unit j, and the active power output of energy storage device j, state variables are constructed; based on the control commands of the synchronous machine, new energy unit, and energy storage device, control variables are constructed; based on the disturbance amount of the system's active power change, disturbance variables are constructed.

[0101] Based on the state variables, control variables, and disturbance variables, construct the discrete state-space model of the system: x(k+1)=Ax(k)+Bu(k)+D d d(k);

[0102] Where x(k), u(k), and d(k) are the system's state variable, control variable, and disturbance variable, respectively; A, B, and D dThese are the system's state matrix, input matrix, and disturbance matrix, respectively.

[0103] In this embodiment, the system's state variable x(k), control variable u(k), and disturbance variable d(k) are respectively:

[0104]

[0105]

[0106] d(k)=ΔP d (k);

[0107] The system's state matrix A, input matrix B, and disturbance matrix D d They are respectively:

[0108]

[0109]

[0110] The disturbance matrix is ​​expanded into a prediction model for N time periods as the system prediction state model:

[0111] X = S x x(0)+S u U+S d D d ;

[0112] Where X is the state variable matrix, x(0) is the first element of the state variable matrix; U is the control command matrix, D d Let X, U, and D be the perturbation variable matrix. d Both are extensions of the original system state vector at N time points; S x S u and S d Let be the extended matrices of the system's state matrix, input matrix, and disturbance matrix at N time points.

[0113] In this embodiment, the state variable matrix is ​​X = [x(0), x(1), ..., x(N)] T The control command matrix is ​​U = [u(0), u(1), ..., u(N-1)] T The perturbation variable matrix is ​​D d =[d(0),d(1),…,d(N-1)] T I is an n-dimensional identity matrix.

[0114]

[0115]

[0116] Step S104: To minimize the objective function of secondary frequency regulation and to establish a system frequency optimization control model with constraints from the system predictive state model, the output constraints of the synchronous machine, the output constraints of the renewable energy unit, and the output constraints of the energy storage system;

[0117] In this embodiment, a system frequency optimization control model is established with the objective function of minimizing the secondary frequency regulation as the constraint condition, and the output constraints of the system predictive state model, the synchronous machine, the renewable energy unit, and the energy storage system. Specifically:

[0118]

[0119] stX=S x x(0)+S u U+S d D d

[0120]

[0121]

[0122]

[0123] Among them, J SFC Let be the objective function for optimizing the second frequency modulation. It is the input control command of the synchronizer i. and These are the upper and lower limits of the control commands for synchronizer i. This is the control command for the new energy unit J. and These are the upper and lower limits of the control commands for the new energy generating unit j. These are the control commands for energy storage device j. and These represent the upper and lower limits of the control commands for energy storage device j, respectively. The upper and lower bounds of the output constraint are subtracted from the output estimate of the power supply in the primary frequency regulation, and P is the set of all inverter interface power supplies (new energy units and energy storage devices).

[0124] In this embodiment, the upper and lower bounds of the output constraint minus the output estimate of the power supply in the first frequency modulation is calculated from the virtual inertia coefficient and the droop control parameters, specifically:

[0125]

[0126] in, and These are the virtual inertia coefficient and droop control parameters for the distributed resource cluster, respectively. Δf(k) and Δf(k) are the measured values ​​of the frequency change rate and frequency deviation at time k, respectively.

[0127] In this embodiment, the frequency change rate and frequency deviation measurements at time k are obtained by the system's wide-area measurement device.

[0128] In this embodiment, the MPC controller optimizes the optimal control command matrix U in each cycle. Although U includes the control action vector u(k) from k=0 to N-1, only the control action vector x(0) of the first cycle is adopted. Vector x(0) includes the secondary frequency regulation active power command of each controllable power source in the system, and then frequency control is achieved by adjusting the active power.

[0129] Step S105: Solve the system frequency optimization control model using a distributed optimization algorithm to make the optimization objective function meet the preset requirements and obtain the optimal decision variable as the frequency regulation result.

[0130] In this embodiment, the system frequency optimization control model is solved using a distributed optimization algorithm, so that the optimization objective function meets the preset requirements, and the optimal decision variable is obtained as the frequency regulation result. Specifically:

[0131] The system frequency optimization control model is transformed into:

[0132]

[0133] stA eq,i X i =b eq,i

[0134] g i (X i )≤0;

[0135] The optimization problem is divided into m subproblems, J i Let X be the cost function of subproblem i. i Let X be the decision variable for subproblem i. i Includes state variables, control commands, and disturbance variables; A eq,i b eq,i These correspond to the linear equality constraints and constant control variables in the system frequency optimization control model, respectively; g i (X i ) represents the expression for the inequality constraints, corresponding to the constraints on all power supply outputs in the system frequency optimization control model.

[0136] Based on the preset iteration step size constant, update the intermediate variable of power supply i in the nth iteration. Control all power sources to interact with their adjacent power sources one by one with their updated intermediate variables; n is the number of iterations;

[0137] According to the preset consensus update rule of distributed optimization, the cost function in the (n+1)th iteration is calculated, and it is determined whether the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition. If so, the current decision variable is used as the frequency regulation result.

[0138] If not, the iteration count is incremented by one until the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition.

[0139] In this embodiment, the intermediate variable of power supply i in the nth iteration is updated according to the preset iteration step size constant. Specifically:

[0140] intermediate variables of subproblems The update expression is:

[0141]

[0142]

[0143] Where I is an n-dimensional identity matrix, For the subproblem equality constraint matrix A eq,i The auxiliary matrix is ​​formed; γ1, γ3, and γ4 are positive iteration step size constants. It is a positive non-decreasing step size constant, where γ0>0, r∈(0,1]; n s The index represents the iteration number; Let be the partial derivative of the objective function with respect to the decision variables in the nth iteration. The expression is:

[0144]

[0145] Ω i =g i,l (X i )≤0,l=1,2,L,N in,i ;

[0146] The inequality constraint vector is expressed as: Let l be the number of inequality constraints in subproblem i. The partial derivative expression of H above is based on the optimization inequality constraint g. i,l Whether it is true or not takes different values.

[0147] In this embodiment, the cost function in the (n+1)th iteration is calculated according to the preset distributed optimization consistency update rule, specifically as follows:

[0148] The distributed optimization consistency update rule is as follows:

[0149] N i ρ represents the set of other power sources adjacent to power source i represented by subproblem i. n =ρ1 / (n+ρ2) is a positive monotonically increasing parameter, and ρ1,ρ2>0, sgn() represents the sign function. ij The information exchange weight between adjacent power sources is defined as:

[0150]

[0151] Where n i n j This indicates the number of adjacent devices for power supply i and power supply j.

[0152] In this embodiment, determining whether the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition specifically involves:

[0153] The stopping update rule for distributed optimization algorithms is:

[0154]

[0155] Where k is the current iteration number, π is the batch number of the error function, Π is the number of batches for calculating the error function, b is the number of cost functions per batch, ε is the limit for error convergence, and ||.||2 represents the 2-function, i.e., the square root of the sum of the squares of all elements of the vector. If the sum of the error functions of Π consecutive batches is less than the limit for error convergence, the algorithm stops converging; if the convergence stopping condition is not met, the iteration number is incremented by one, i.e., n = k + 1, and the intermediate variable of power source i in the nth iteration is updated; all power sources are controlled to interact with their adjacent power sources one by one with their updated intermediate variables; according to the preset consistency update rule of distributed optimization, the cost function in the (n+1)th iteration is calculated until the cost function in the (n+1)th iteration meets the preset convergence stopping condition.

[0156] Implementing the embodiments of the present invention has the following effects:

[0157] The power grid frequency regulation method proposed in this invention uses the sum of a second cost function from the current cycle to a preset future cycle as the optimization objective function for secondary frequency regulation, and performs frequency optimization control by minimizing the optimization objective function of secondary frequency regulation, thereby reducing the system operating cost during the secondary frequency regulation stage. Furthermore, by constructing a system prediction state model using state variables, control variables, and disturbance variables, it reduces errors in renewable energy prediction data, ensures the robustness of frequency control, improves the accuracy of power grid frequency control, and reduces the impact of prediction errors and random disturbances on the system frequency regulation performance. In addition, this invention uses a distributed optimization algorithm to solve the system frequency optimization control model, reducing the computational complexity and communication burden of real-time frequency control.

[0158] Example 2

[0159] Please refer to Figure 2 The present invention provides a power grid frequency regulation device, comprising: an acquisition module 201, an objective function calculation module 202, a state prediction module 203, an optimization control module 204, and a regulation result output module 205.

[0160] The acquisition module 201 is used to acquire load and power frequency regulation reserve data, power model information, load data and power cost coefficient of the power grid system.

[0161] The objective function calculation module 202 is used to calculate the first cost function of the active power regulation of the distributed resource cluster based on the power cost coefficient and the reference output active power of the distributed resource cluster; and to calculate the sum of the second cost functions from the current period to the future preset period as the optimization objective function of the secondary frequency regulation based on the first cost function, the frequency regulation mileage cost and the optimization weight coefficient of the frequency deviation.

[0162] The state prediction module 203 is used to construct a frequency response model of the power grid system based on the load and power supply frequency regulation reserve data, the power supply model information and the load data, generate state variables, control variables and disturbance variables, and construct a system prediction state model.

[0163] The optimization control module 204 is used to establish a system frequency optimization control model with the objective function of secondary frequency regulation being minimized and with the system predictive state model, the output constraints of the synchronous machine, the output constraints of the renewable energy unit and the output constraints of the energy storage system as constraints.

[0164] The regulation result output module 205 is used to solve the system frequency optimization control model through a distributed optimization solution algorithm, so that the optimization objective function reaches the preset requirements and obtains the optimal decision variable as the frequency regulation result.

[0165] The state prediction module 203 includes: a secondary frequency modulation model construction unit, a frequency response model construction unit, and a system prediction state model construction unit;

[0166] The secondary frequency regulation model construction unit is used to generate upper and lower limits for the output of various power sources in the power grid system based on load and power source frequency regulation reserve data and power source model information; based on the upper and lower limits, it calculates the active power output value of each type of power source, and uses the active power output value of all the power sources as the secondary frequency regulation model of the distributed resource cluster, specifically:

[0167] Based on load and power frequency regulation reserve data and power model information, upper and lower limits of control commands for synchronous machines, new energy units and energy storage devices in the power grid system are generated respectively.

[0168] Based on the dynamic adjustment time constant, control command, and upper and lower limits of the control command for various power sources, the active power output values ​​of the synchronous machine, new energy units, and energy storage devices in the power grid system are generated respectively.

[0169] The upper and lower limits of the control commands of the energy storage device are calculated based on the charging and discharging efficiency of the energy storage device and the lower and upper limits of the controllable adjustable amount of the energy storage device.

[0170] The active power output values ​​of synchronous machines, new energy generating units, and energy storage devices in the power grid system are used as the secondary frequency regulation model of the distributed resource cluster.

[0171] The frequency response model construction unit is used to construct a frequency response model of the power grid system based on the load data and the secondary frequency regulation model, specifically as follows:

[0172] Based on the active power output values ​​of synchronous machines, new energy generating units, and energy storage devices, load data, and the oscillation equation of the power grid system, a frequency response model of the power grid system is constructed:

[0173]

[0174] Where G is the set of synchronous machines, ΔP i SG R is the active power output value of synchronous machine i, and R is the set of new energy generating units. Here, E represents the active power output of the new energy unit j, and E represents the collection of energy storage devices. Where M is the active power output of energy storage device j, D is the overall system inertia, and P is the overall system damping coefficient. L Here, Δf represents the system load data, ΔP represents the frequency deviation of the power grid system, and ΔP represents the frequency deviation of the power grid system. d It is the disturbance quantity of the system's active power change.

[0175] The system predictive state model construction unit is used to generate state variables, control variables, and disturbance variables based on the frequency response model and the second-order frequency modulation model, and to construct the system predictive state model, specifically as follows:

[0176] Based on the frequency deviation of the power grid system, the active power output of synchronous machine i, the active power output of new energy unit j, and the active power output of energy storage device j, state variables are constructed; based on the control commands of the synchronous machine, new energy unit, and energy storage device, control variables are constructed; based on the disturbance amount of the system's active power change, disturbance variables are constructed.

[0177] Based on the state variables, control variables, and disturbance variables, construct the discrete state-space model of the system: x(k+1)=Ax(k)+Bu(k)+D d d(k);

[0178] Where x(k), u(k), and d(k) are the system's state variable, control variable, and disturbance variable, respectively; A, B, and D d These are the system's state matrix, input matrix, and disturbance matrix, respectively.

[0179] The disturbance matrix is ​​expanded into a prediction model for N time periods as the system prediction state model:

[0180] X = S x x(0)+S u U+S d D d ;

[0181] Where X is the state variable matrix, x(0) is the first element of the state variable matrix; U is the control command matrix, D d Let X, U, and D be the perturbation variable matrix. d Both are extensions of the original system state vector at N time points; S x S u and S d Let be the extended matrices of the system's state matrix, input matrix, and disturbance matrix at N time points.

[0182] The optimization control module 204 includes: a system frequency optimization control model construction unit;

[0183] The system frequency optimization control model construction unit is used to establish a system frequency optimization control model with the objective function of minimizing secondary frequency regulation and with constraints from the system predictive state model, the output constraints of the synchronous machine, the output constraints of the renewable energy unit, and the output constraints of the energy storage system. Specifically:

[0184]

[0185] stX=Sx x(0)+S u U+S d D d

[0186]

[0187]

[0188]

[0189] Among them, J SFC Let be the objective function for optimizing the second frequency modulation. It is the input control command of the synchronizer i. and These are the upper and lower limits of the control commands for synchronizer i. This is the control command for the new energy unit J. and These are the upper and lower limits of the control commands for the new energy generating unit j. These are the control commands for energy storage device j. and These represent the upper and lower limits of the control commands for energy storage device j, respectively. The upper and lower bounds of the output constraint are subtracted from the output estimate of the power supply in the primary frequency regulation, and P is the set of all inverter interface power supplies (new energy units and energy storage devices).

[0190] The upper and lower bounds of the output constraints minus the output estimate of the power supply in primary frequency regulation are calculated from the virtual inertia coefficient and the droop control parameters, specifically:

[0191]

[0192] in, and These are the virtual inertia coefficient and droop control parameters for the distributed resource cluster, respectively. Δf(k) and Δf(k) are the measured values ​​of the frequency change rate and frequency deviation at time k, respectively.

[0193] The regulation result output module 205 includes: a solution unit;

[0194] The solution unit is used to solve the system frequency optimization control model using a distributed optimization algorithm, so that the optimization objective function reaches the preset requirements and the optimal decision variables are obtained as the frequency regulation result, specifically:

[0195] The system frequency optimization control model is transformed into:

[0196]

[0197] stAeq,i X i =b eq,i

[0198] g i (X i )≤0;

[0199] The optimization problem is divided into m subproblems, J i Let X be the cost function of subproblem i. i Let X be the decision variable for subproblem i. i Includes state variables, control commands, and disturbance variables; A eq,i b eq,i These correspond to the linear equality constraints and constant control variables in the system frequency optimization control model, respectively; g i (X i ) represents the expression for the inequality constraints, corresponding to the constraints on all power supply outputs in the system frequency optimization control model.

[0200] Based on the preset iteration step size constant, update the intermediate variable of power supply i in the nth iteration. Control all power sources to interact with their adjacent power sources one by one with their updated intermediate variables; n is the number of iterations;

[0201] According to the preset consensus update rule of distributed optimization, the cost function in the (n+1)th iteration is calculated, and it is determined whether the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition. If so, the current decision variable is used as the frequency regulation result.

[0202] If not, the iteration count is incremented by one until the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition.

[0203] The aforementioned power grid frequency regulation device can implement a power grid frequency regulation method according to the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining contents of this application embodiment can be referred to the contents of the above method embodiments, and will not be repeated in this embodiment.

[0204] Implementing the embodiments of the present invention has the following effects:

[0205] The objective function calculation module of this invention uses the sum of the second cost functions from the current cycle to a future preset cycle as the optimization objective function for secondary frequency regulation. The optimization control module performs frequency optimization control by minimizing the optimization objective function of secondary frequency regulation, reducing the system operating cost during the secondary frequency regulation stage. The state prediction module constructs a system prediction state model using state variables, control variables, and disturbance variables, reducing errors in new energy prediction data, ensuring the robustness of frequency control, improving the accuracy of grid frequency control, and reducing the impact of prediction errors and random disturbances on the system frequency regulation performance. Furthermore, the control result output module of this invention solves the system frequency optimization control model using a distributed optimization algorithm, reducing the computational complexity and communication burden of real-time frequency control.

[0206] Example 3

[0207] Accordingly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform a power grid frequency regulation method as described in any of the above embodiments.

[0208] For example, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0209] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0210] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0211] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function, etc.; the data storage area may store data created based on the use of the mobile terminal, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0212] Wherein, if the modules / units integrated in the terminal device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, it can implement the steps of the various method embodiments described above. Wherein, the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0213] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for power grid frequency regulation, characterized in that, include: Acquire load and power frequency regulation reserve data, power model information, load data and power cost coefficient of the power grid system; Based on the power cost coefficient and the reference output active power of the distributed resource cluster, a first cost function for the active power regulation of the distributed resource cluster is calculated; based on the first cost function, the frequency regulation mileage cost, and the optimization weight coefficient of the frequency deviation, the sum of the second cost functions from the current period to the future preset period is calculated as the optimization objective function of the secondary frequency regulation. Based on the load and power supply frequency regulation reserve data, the power supply model information, and the load data, a frequency response model of the power grid system is constructed, state variables, control variables, and disturbance variables are generated, and a system predictive state model is constructed. A system frequency optimization control model is established with the objective function of minimizing the secondary frequency regulation as the constraint condition, and with the output constraints of the system predictive state model, the synchronous machine, the renewable energy unit and the energy storage system as the constraints. The system frequency optimization control model is solved by a distributed optimization algorithm so that the optimization objective function meets the preset requirements and the optimal decision variable is obtained as the frequency regulation result. The first cost function of the active power adjustment of the distributed resource cluster is expressed as: ； Where, ∆P i (k) represents the active power adjustment of resource cluster i at time k. For ∆P i The first cost function of (k), P i (k) represents the reference output active power of distributed resource cluster i in the scheduling plan at time k, a i and b i These are the primary and secondary scheduling cost coefficients for the distributed resource cluster, respectively. The sum of the second cost functions from the current period to a preset future period is used as the optimization objective function for the second frequency modulation: ； ； Among them, J SFC To optimize the objective function, G is the set of all synchronous machines i in the system, P is the set of all power electronic interface asynchronous power supplies j in the system, and q f Here, ΔL represents the optimized weighting coefficient of the frequency deviation term Δf(k), where N represents the next N periods. i For FM mileage cost, and These represent the secondary frequency modulation commands at time k and time k-1, respectively. i This is the FM mileage cost coefficient.

2. The method for power grid frequency regulation as described in claim 1, characterized in that, The process involves constructing a frequency response model of the power grid system based on the load and power supply frequency regulation reserve data, the power supply model information, and the load data; generating state variables, control variables, and disturbance variables; and constructing a system predictive state model. Specifically: Based on load and power supply frequency regulation reserve data and power supply model information, the upper and lower limits of the output of various power sources in the power grid system are generated; based on the upper and lower limits of the output, the active power output value of each of the power sources is calculated, and the active power output value of all the power sources is used as the secondary frequency regulation model of the distributed resource cluster. Based on the load data and the secondary frequency regulation model, a frequency response model of the power grid system is constructed; Based on the frequency response model and the second frequency modulation model, state variables, control variables, and disturbance variables are generated, and a system prediction state model is constructed.

3. The method for power grid frequency regulation as described in claim 2, characterized in that, The process involves generating upper and lower limits for the output of various power sources in the power grid system based on load and power source frequency regulation reserve data and power source model information. Based on these limits, the active power output values ​​of each power source are calculated. Finally, the active power output values ​​of all power sources are used as the secondary frequency regulation model for the distributed resource cluster. Specifically: Based on load and power frequency regulation reserve data and power model information, upper and lower limits of control commands for synchronous machines, new energy units and energy storage devices in the power grid system are generated respectively. Based on the dynamic adjustment time constant, control command, and upper and lower limits of the control command for various power sources, the active power output values ​​of the synchronous machine, new energy units, and energy storage devices in the power grid system are generated respectively. The upper and lower limits of the control commands of the energy storage device are calculated based on the charging and discharging efficiency of the energy storage device and the lower and upper limits of the controllable adjustable amount of the energy storage device. The active power output values ​​of synchronous machines, new energy generating units, and energy storage devices in the power grid system are used as the secondary frequency regulation model of the distributed resource cluster.

4. The method for power grid frequency regulation as described in claim 3, characterized in that, The step of constructing a frequency response model for the power grid system based on the load data and the secondary frequency regulation model is as follows: Based on the active power output values ​​of synchronous machines, new energy generating units, and energy storage devices, load data, and the oscillation equation of the power grid system, a frequency response model of the power grid system is constructed: ； Where G is the set of synchronous machines, R is the active power output value of synchronous machine i, and R is the set of new energy generating units. Here, E represents the active power output of the new energy unit j, and E represents the collection of energy storage devices. Where M is the active power output of energy storage device j, D is the overall system inertia, and P is the overall system damping coefficient. L Here, Δf represents the system load data, ΔP represents the frequency deviation of the power grid system, and ΔP represents the frequency deviation of the power grid system. d It is the disturbance quantity of the system's active power change.

5. The method for power grid frequency regulation as described in claim 4, characterized in that, The step of generating state variables, control variables, and disturbance variables based on the frequency response model and the second-order frequency modulation model, and constructing a system prediction state model, specifically involves: Based on the frequency deviation of the power grid system, the active power output of synchronous machine i, the active power output of new energy unit j, and the active power output of energy storage device j, state variables are constructed; based on the control commands of the synchronous machine, new energy unit, and energy storage device, control variables are constructed. Based on the disturbance amount of the system's active power change, construct the disturbance variable; Based on the state variables, control variables, and disturbance variables, construct the discrete state-space model of the system: ; Where x(k), u(k), and d(k) are the system's state variable, control variable, and disturbance variable, respectively; A, B, and D d These are the system's state matrix, input matrix, and disturbance matrix, respectively. The disturbance matrix is ​​expanded into a prediction model for N time periods as the system prediction state model: ； Where X is the state variable matrix, x(0) is the first element of the state variable matrix; U is the control command matrix, D d Let X, U, and D be the perturbation variable matrix. d Both are extension vectors of the original system state vector at N time points; , and Let be the extended matrices of the system's state matrix, input matrix, and disturbance matrix at N time points.

6. The method for power grid frequency regulation as described in claim 5, characterized in that, The system frequency optimization control model is established by minimizing the objective function of secondary frequency regulation and taking the output constraints of the system predictive state model, the synchronous machine, the renewable energy unit, and the energy storage system as constraints. Specifically: ； Among them, J SFC Let be the objective function for optimizing the second frequency modulation. It is the input control command of the synchronizer i. and These are the upper and lower limits of the control commands for synchronizer i. This is the control command for the new energy unit J. and These are the upper and lower limits of the control commands for the new energy generating unit j. These are the control commands for energy storage device j. and These represent the upper and lower limits of the control commands for energy storage device j, respectively. The upper and lower bounds of the output constraints are subtracted from the estimated output value of the power supply in the primary frequency regulation, and P is the set of all new energy units and energy storage devices.

7. The method for power grid frequency regulation as described in claim 6, characterized in that, The upper and lower bounds of the output constraint minus the output estimate of the power supply in primary frequency regulation are calculated from the virtual inertia coefficient and the droop control parameters, specifically: ； in, These are the virtual inertia coefficient and droop control parameters for the distributed resource cluster, respectively. and These are the frequency change rate and frequency deviation measurements at time k, respectively.

8. The method for power grid frequency regulation as described in claim 7, characterized in that, The process involves solving the system frequency optimization control model using a distributed optimization algorithm, ensuring the optimization objective function meets preset requirements, and obtaining the optimal decision variables as the frequency regulation result. Specifically: The system frequency optimization control model is transformed into: ； The optimization problem is divided into m subproblems, J i Let X be the cost function of subproblem i. i Let X be the decision variable for subproblem i. i Includes state variables, control commands, and disturbance variables; A eq,i b eq,i These correspond to the linear equality constraints and constant control variables in the system frequency optimization control model, respectively; g i (X) i ) is the expression for the inequality constraint, corresponding to the constraint conditions of all power supply outputs in the system frequency optimization control model; Based on the preset iteration step size constant, update the intermediate variable of power supply i in the nth iteration. Control all power sources to interact with their adjacent power sources one by one using their updated intermediate variables; n is the number of iterations; According to the preset consensus update rule of distributed optimization, the cost function in the (n+1)th iteration is calculated, and it is determined whether the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition. If so, the current decision variable is used as the frequency regulation result. If not, the iteration count is incremented by one until the cost function in the (n+1)th iteration satisfies the preset convergence stopping condition.

9. A device for regulating power grid frequency, characterized in that, The power grid frequency regulation method as described in any one of claims 1 to 8 is adopted; The power grid frequency regulation device includes: an acquisition module, an objective function calculation module, a state prediction module, an optimization control module, and a regulation result output module; The acquisition module is used to acquire load and power frequency regulation reserve data, power model information, load data and power cost coefficient of the power grid system. The objective function calculation module is used to calculate the first cost function of the active power regulation of the distributed resource cluster based on the power cost coefficient and the reference output active power of the distributed resource cluster; and to calculate the sum of the second cost functions from the current period to the future preset period based on the first cost function, the frequency regulation mileage cost and the optimization weight coefficient of the frequency deviation as the optimization objective function of the secondary frequency regulation. The state prediction module is used to construct a frequency response model of the power grid system based on the load and power supply frequency regulation reserve data, the power supply model information and the load data, generate state variables, control variables and disturbance variables, and construct a system predicted state model. The optimization control module is used to establish a system frequency optimization control model with the objective function of secondary frequency regulation being minimized and with constraints such as the system predictive state model, the output constraints of the synchronous machine, the output constraints of the renewable energy unit, and the output constraints of the energy storage system. The regulation result output module is used to solve the system frequency optimization control model through a distributed optimization solution algorithm, so that the optimization objective function reaches the preset requirements and the optimal decision variable is obtained as the frequency regulation result.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program; wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform a power grid frequency regulation method as described in any one of claims 1 to 8.

Images